Cause for Alarm
Addressing the Alarm-Setting Process in Real-Time Detection Systems Implementation
Working from Home but Missing Your Synergist? Update Your Address
If you’ve been working from home during the pandemic, please consider updating your address with AIHA. You can change your address by editing your profile through AIHA.org. To ensure uninterrupted delivery of The Synergist, designate your home address as “preferred” on your profile. Update your address now.
In occupational and environmental health and safety, the use of real-time detection systems (RTDS), also referred to as direct-reading instruments (DRI), has increased significantly in recent years. This increased usage has introduced many OEHS professionals to the difficulties of RTDS implementation, which Spencer Pizzani, a member of AIHA’s Real-Time Detection Systems Committee (RTDSC), described in an article that appeared in the November 2022 Synergist. One of these difficulties is aligning alarms coming from detected conditions to specific actions by all consumers or stakeholders of the data. The authors of this article, who have experience with the manufacturing, research and teaching, and field use of real-time detection systems, have also written a white paper on this topic, “Establishing a Process for Setting Real-Time Detection System Alarms,” which AIHA recently published online (PDF). Here, we introduce a five-step process for establishing an alarm setpoint.
BACKGROUND AIHA previously addressed the monitoring of hazardous agents in a 2009 joint publication with the Center for Chemical Process Safety (CCPS), Continuous Monitoring for Hazardous Material Releases. The topic was further explored by the United Kingdom’s Health and Safety Executive in “Review of Alarm Setting for Toxic Gas and Oxygen Detectors.” Neither document satisfactorily presented a rational process to guide users about how to arrive at an appropriate alarm setpoint.
In 2016, AIHA organized a “sensor summit” that brought together a variety of disciplines and resulted in the report “The Future of Sensors.” Three years later, burgeoning deployment of sensors in technology and processes (such as the “Internet of Things”) pushed the AIHA Content Portfolio Advisory Group, which advises the AIHA Board of Directors on content development, to merge two previously separate priorities into one: Big Data and sensor technology.
Also in 2019, the RTDSC joined forces with the U.S. Department of Energy-sponsored Energy Facilities Contractors Operating Group to address the manner of interpretation of peak concentration data recorded by four-gas monitors versus established occupational health ceiling limits. The document produced by that collaboration did not include an alarm-setting process but alluded to its complexities, which were discussed in Pizzani’s Synergist article.
In 2020, the RTDSC proposed to the AIHA Board a project that would “establish a vetted alarm settings process and include case study examples as demonstration of the process validity across a wide variety of instruments used for real-time detection of hazards and assessment of exposures.” Over the past two years, we created the five-step process presented in Table 1.
Table 1. Proposed Steps for an Alarm-Setting Process
With the process defined, attention turned to a specific problem: how should one go about setting an alarm threshold for action?
The use of sensing technology for hazard detection is not new. In 1930, Swiss physicist Walter Jaeger discovered that his prototype poison gas detector responded to smoke particles (although it took more than 30 years to refine a commercial smoke detector for home use, according to the U.S. Nuclear Regulatory Commission). Such detectors were purely qualitative; any smoke would result in an alarm. And despite many decades of sensor development for the detection of hazards, a parallel effort to ensure clear communication of the sensor output to result in specific actions did not materialize.
The International Society of Automation (ISA) invested a significant amount of effort on the alarm topic, which resulted in the 2009 release of ISA 18.2, Management of Alarm Systems for the Process Industries. Development of the ISA 18.2 standard was motivated by the connection between poor alarm management and process safety accidents, such as the disastrous toxic chemical release that occurred in Bhopal, India, in 1984; the oil storage facility fires in Milford Haven and Buncefield, U.K., in 1983 and 2005, respectively; and the 2005 refinery explosion in Texas City, Texas. ISA developed a lifecycle model for successful alarm system design, implementation, operation, and management to address the observation in a 1998 report commissioned by HSE that “poor performance costs money in lost production and plant damage and weakens a very important line of defense against hazards to people.” An overview of the ISA 18.2 process appears in Figure 1.
From a series of technical reports supporting ISA 18.2, it is clear that extensive consideration of an alarm management philosophy is needed to successfully deploy sensors. But ISA 18.2 identifies “alarm setpoint determination” as only a recommended, not a required, element of that philosophy, and the technical reports referenced in the standard do not provide any guidance as to how to fully accomplish this.
While the ISA alarm management philosophy was primarily designed for facility process sensors and alarms in production and manufacturing plants, it serves the OEHS community well for fixed and mobile (including wearable) sensors. However, the ISA standard still lacks any guidance on all the considerations that should be formally analyzed in the determination of an alarm threshold value. The recent AIHA white paper presents an initial approach.
Figure 1. The alarm management lifecycle as described in ISA 18.2, Management of Alarm Systems for the Process Industries. Adapted from Exida.com, “The ISA-18.2 / IEC 62682 Alarm Management Lifecycle” (PDF).
A PROPOSED ALARM-SETTING PROCESS The AIHA white paper proposes a sequence of five steps for setting alarms. These steps will allow a defensible selection of sensing technology that is fit for purpose in the context of use of the sensor, instrument, or device to detect and alarm to the hazard or condition. Briefly, the process entails:
Step 1: Statement of purpose in target of condition to detect. The first step should be a succinct but complete analysis of the goal of detecting a hazard or condition. The successful completion of the analysis is important for the subsequent steps, determination of the selected technology, and the alarm thresholds, and it can help OEHS professionals clarify what is not intended as a goal to other stakeholders such as workers, management, regulators, or the public. While developing the white paper, we discussed the following scenarios:
• controlling workers’ entry to hazardous atmospheres • intervening in episodes of excessive drivers’ fatigue • alerting workers to rising chlorine dioxide levels • managing ventilation efficiency through aerosol monitoring (particle counting) and carbon dioxide concentration monitoring
Step 2: Selection of objectives (parameters) around the condition to detect. The second step should be the analysis of considerations for alarm setting. Setting an alarm may serve as a safety monitoring warning of imminent conditions such as fire, explosion, or asphyxia, or provide indication of health monitoring for more chronic conditions such as exposure levels associated with carcinogens or other chronic health consequences. A list of potential objectives adapted from the 2013 HSE publication appears in the sidebar below.
Step 3: Selection of the device and determination that it is fit for purpose. The third step entails rigorous evaluation of the context for use of the sensor, instrument, or device. Performance characteristics can vary widely between and within instrument classes, makes and models, and sensors. These characteristics must be considered in relation to critical factors such as the types of agents detected, the limit of quantification for a specific agent, and accuracy of detection. An instrument must be fit for purpose for both the identified context and objective of its use. It is possible to select an instrument that can detect a specific agent but is still not fit for purpose for numerous reasons. For example, a field condition may foul the sensor (as in the case of catalytic sensor poisons) or bias the results (for example, humidity for a photoionization detector).
Five specific considerations should guide the evaluation:
The target of the monitoring. What is the hazard or condition that must be recognized, qualitatively or quantitatively, by the sensor?
The operability of the instrument in the environment where it will be deployed. Of particular importance is the confirmation that communication of the condition, and expected actions by those receiving it, will work seamlessly. This may entail training those expected to respond to the alarm and requiring them to demonstrate proficiency.
The environmental conditions to be encountered in the field. Once deployed, will the sensor work in all conditions of temperature and humidity it encounters? Will wireless technology be available to support communication?
Human factors. The planned alarm must alert the user to a condition and possibly direct further assessment (to investigate, for example) or action (such as vacating a confined space). The alarm must be timely to allow sufficient time for execution of a defined response. Consideration must be given to the human capabilities and limitations that will be in place when the alarm occurs; that is, the limitations of the response must be considered.
Specific instrument certification requirements. When assessing an instrument’s fitness for purpose, a user must consider and verify that the planned use is consistent with the instrument’s certifications. These may include safety performance certifications, such as those for devices that are acceptable for use in explosive atmospheres, and response performance certifications, which indicate that devices are acceptable for use at specific sites, such as mining operations.
The process of evaluating instruments and systems is a complex task that should be approached carefully, especially for unusual agents or conditions. In 2017, the RTDSC developed a Standardized Equipment Specification Sheet (SESS), which can assist OEHS professionals in identifying important performance characteristics when determining whether instruments are correct for their intended purpose. Other committee-authored documents on RTDS include “Reporting Specifications for Electronic Real-Time Gas and Vapor Detection Equipment,” “Competency Framework: A Prime Resource for Using DRIs in Gas Monitoring and Detection,” and “Guidance on the Use of Direct-Reading Instruments.”
Step 4: Establish the alarm setpoints and confirm the alarm response process. As defined in ISA TR18.2.1-2018, Alarm Philosophy, an “alarm setpoint” is a user-configurable threshold value that triggers an annunciation for one or more signals. In this article and in the recently published white paper, the terms “limits,” “levels,” or “criteria threshold,” as they relate to setting an alarm, are all synonymous with “setpoint.”
The following progression of questions will help the user construct a successful defense of the setpoint and, in turn, facilitate successful instrument deployment:
“Do we have a regulatory limit?” If the instrument is being used for the purpose of demonstrating compliance to a regulation, then the limit set by the regulation is the user’s starting point for selecting an alarm setpoint.
“Do we have a limit assigned by contract, corporate policy, or other authoritative source?” If the purpose of the instrument is to support conformance to the provisions of a contract, meet corporate expectations, or fulfill requirements for some other documented, rational, and thoroughly communicated reason, the criterion prescribed by this agreement may be used as the alarm setpoint. This type of setpoint also encompasses limits recommended by organizations such as ASTM, NIOSH (for the agency’s Recommended Exposure Limits), ACGIH (for Threshold Limit Values), or AIHA (for Workplace Environmental Exposure Levels).
“Does the setpoint need to be selected based on field criteria?” Such criteria may be subjective (such as ergonomic, fatigue, or odor parameters) or based on variable but objective criteria (such as heat strain). The setpoint may also be selected based on multiple sources, such as infrared radiation and ambient temperature and the worker’s encapsulating personal protective equipment, or it may depend on the type or phase of hazardous material present. The process of assigning a setpoint should be evaluated by the Tier 4 (that is, advanced) user deploying the alarm technology, with Tier 1 or 2 (basic or intermediate) users validating that they are able to make the required determination for the correct limit. (For a discussion of the different tiers of users of RTDS, refer to AIHA’s “Technical Framework: Guidance on Use of Direct-Reading Instruments.”)
“How may we implement the precautionary principle when nothing else exists to guide us on establishing a setpoint?” When no authoritative criteria to support assigning an alarm setpoint exists, a threshold percentage of a limit as determined by health and safety professionals or supporting technical opinions—from the employer’s legal, toxicology, or product stewardship departments, for example—should be presented to stakeholders for consideration and adoption. This is an example of the precautionary principle.
“Is the chosen setpoint a product of understanding the variability of historic data?” On occasion, alarm setpoints can be anticipated from historical data, and adjustments may be made without incurring risks of adverse consequences to people or property.
Step 5: Communication of the alarm setpoints and supporting data. The alarm-setting process is complete when the selected instruments’ alarm setpoints and datalogging configurations, and the interpretation of the data, are recorded and shared with stakeholders. The white paper contains a template for documenting each step in the alarm-setting process. It is also important for the alarm-setting process to be reviewed on a predetermined basis to ensure the device and its alarm configurations remain fit for purpose. Review is also necessary when a change in the field-deployed conditions occurs, a practice often referred to as management of change in the process safety industry.
We hope that this article will prompt readers to examine the white paper (PDF) and to test it through use.
AIHce EXP 2023 in Phoenix will include eight education sessions and three professional development courses addressing real-time detection systems. The presenters in these sessions are RTDSC members with active engagement and experience in deploying real-time instruments. We welcome constructive criticism of the alarm-setting process and stories reporting the efforts of professionals who attempt its implementation. We and the members of the RTDSC are committed to advancing the knowledgeable adoption of sensor technologies and exploring the utility of Big Data.
EVISON CAREFOOT, CIH, PEng, CRSP, is senior consultant with Salus Services Limited.
EMANUELE CAUDA, PhD, is the director of the NIOSH Center for Direct Reading and Sensor Technologies as well as an adjunct professor at the University of Pittsburgh Graduate School of Public Health.
ROBERT HENDERSON is past chair of the AIHA Real-Time Detection Systems Committee and president of GfG Instrumentation Inc.
STEVEN JAHN, CIH, MBA, FAIHA, is a technical advisor for Savannah River Nuclear Solutions LLC in Aiken, South Carolina, and president of Jahn Industrial Hygiene LLC.
WILLIAM MILLS, PhD, MSc, CIH, CChem, is research officer for the AIHA Real-Time Detection Systems Committee and an associate professor in the College of Engineering and Engineering Technology at Northern Illinois University.
Send feedback to The Synergist.
Alarm-Setting Considerations for Installing Gas Detectors
Based on HSE’s “Review of Alarm Setting for Toxic Gas and Oxygen Detectors,” consider the following when installing gas detectors: •whether detectors should be fixed and whether personnel should be issued with portable, which includes personal, detectors •the location of the fixed detector and whether the work area is occupied or unoccupied •whether the fixed detector is a point detector (also known as a beam detector) or an open-path detector (also known as a line-of-sight detector) •whether egress is difficult or time-consuming if there is an emergency •whether Workplace Environmental Exposure Levels (WEELs), other exposure limit values, or other health-based levels (such as Immediately Dangerous to Life and Health, or IDLH, values) exist •instantaneous or time-weighted average (TWA) alarm •false alarms caused by instrumental effects and interferent gases, which may also be classified as spurious •the characteristics of the sources and the potential rate of gas buildup •time to alarm of the detection system •number of alarm levels (for example, high and low levels) •mixture of gases or vapors •whether the fixed detector is a diffusion-type or a pumped-type (also known as sample draw) detector •whether the monitoring area is in a hazardous or permit-required location, which may introduce additional alarm-setting constraints •whether the RTDS is able to communicate real-time alarms to other individuals (for example, wireless communication between confined space entrants and attendants) •whether RTDS results will be used to coordinate off-site response or third-party rescue •whether RTDS results will be used for instantaneous (peak or ceiling) or time history (STEL or TWA) alarms •whether background variations and events from the process can trigger “spurious” alarms •environmental conditions that may exceed the design or certification limits of the RTDS (for example, temperature, pressure, and humidity) •background conditions that can affect the proper performance of the sensors (such as lack of oxygen due to natural or deliberate inerting of confined spaces) •environmental conditions such as high noise levels that can prevent the user from knowing when an alarm occurs •activities such as hot work that introduce additional hazards •other external or internal factors that may affect or prevent the RTDS from activating the appropriate alarms
AIHA: “Big Data and Sensor Technology Resources.”
AIHA: “Competency Framework: Field Use of Direct-Reading Instruments for Detection of Gases and Vapors” (2020).
AIHA: "Establishing a Process for the Setting of Real-Time Detection System Alarms" (PDF, February 2023).
AIHA: “Reporting Specifications for Electronic Real Time Gas and Vapor Detection Equipment,” version 3 (PDF, 2021).
AIHA: “Technical Framework: Guidance on Use of Direct-Reading Instruments” (2020).
AIHA: “The Future of Sensors: Protecting Worker Health Through Sensor Technologies” (PDF, April 2016).
Center for Chemical Process Safety and AIHA: Continuous Monitoring for Hazardous Material Release (2009).
Department of Energy: “A Practical Guide for Use of Real-Time Detection Systems for Worker Protection and Compliance with Occupational Exposure Limit” (May 2019).
Exida: “Alarm Philosophy Document Template” (PDF, October 2011).
Exida: “Plug the Holes in the Swiss Cheese Model” (2020).
Exida: “The ISA-18.2 / IEC 62682 Alarm Management Lifecycle” (PDF, 2022).
Health and Safety Executive: “Review of Alarm Setting for Toxic Gas and Oxygen Detectors” (PDF, 2013).
Health and Safety Executive: “The Management of Alarm Systems” (PDF, 1998).
International Society of Automation: ANSI/ISA 18.2, Management of Alarm Systems for the Process Industries (2009).
International Society of Automation: ANSI/ISA 18.2-2016, Management of Alarm Systems for the Process Industries (2016).
International Society of Automation: ISA TR18.2.1-2018, Alarm Philosophy (2018).
International Society of Automation: “Understanding and Applying the ANSI/ISA 18.2 Alarm Management Standard” (PDF, 2010).
Nuclear Regulatory Commission: “Backgrounder on Smoke Detectors” (May 2020).
The Synergist: “Five Common Difficulties in Real-Time Detection System Implementation” (November 2022).
SynergistNOW: “The Evolving Role of Sensors to Protect Worker Health” (January 2018).